Pattern Classifications in Cognitive Informatics - Computational Intelligence Paradigms in Advanced Pattern Classification - page 40

So far, the authors of the above analysed image data from medical imaging, and
more specifically, analysed lesions occurring within the central nervous system,
i.e. the spinal cord, in the bones of feet, palms and wrists. The second,
independent type of data analysed consisted in economic and financial figures of
companies, where the aim was to assess whether the financial, strategic and
marketing decisions suggested by the system were reasonable.
Here, the author will concentrate on cognitive systems analysing the meaning
of image data, with particular attention to UBIAS (Understanding Based Image
Analysis Systems) analysing lesions within foot bones. The novelty presented here
is that the UBIAS system will analyse lesions within the bones of the whole foot,
as previously efforts were only made to analyse foot bones excluding phalanx
bones. Including the phalanx bones in the analysis of foot bone lesions makes the
process of understanding the analysed medical (health) situation much richer, so
the process can identify the right lesion more unambiguously, consequently
yielding a deeper cognitive analysis and the semantic analysis as such of the lesion
observed.

2 Semantic Analysis Stages

Semantic analysis is the key to the correct operation of cognitive data analysis
systems. When this analysis is conducted, several different (but equally important
for the analysis) processes occur: interpretation, description, analysis and
reasoning.
The main stages of semantic analysis are as follows [26], [43], [46]:

• data pre-processing:

 filtration and amplification;



approximation;

 coding;

• data presentation:



segmentation;

 recognising picture primitives;

 identifying relations between picture primitives;

• linguistic perception;

• syntactic analysis;

• pattern classification;

•

data classification;
•

feedback;
• cognitive resonance;

• data understanding.

The main stages of semantic analysis are shown in Figure 1. A clear majority of
the above semantic analysis stages deals with the data understanding process, as
from the beginning of the syntactic analysis conducted using the formal grammar